The initial search strategy yielded a total of 1361 references from the online databases, and 500 references were excluded because of duplication. Additionally, 838 references were excluded after screening the titles and abstracts, and 5 references were excluded after the full text was screened. Finally, eighteen studies were enrolled in the meta-analysis. A detailed flowchart of the study selection is shown in Fig. 1. In the pooled analysis, we did not summarize the low-dose groups of phase 2 clinical trials, including erenumab 7 mg/21 mg/28 mg, galcanezumab 5 mg/50 mg, and eptinezumab 10 mg/30 mg; as these low-dose groups had only had only partial efficacy and did not enter phase 3 clinical trials.
Study characteristics and study quality
The eighteen included studies recruited 11,099 patients with migraines. Table 1 presents the characteristics of the included studies. Most of the patients were females (n = 9405; 84.74%), and there were only 1694 males (15.33%). Eighteen studies focused on four CGRP mAbs: erenumab (6), fremanezumab (5), galcanezumab (5), and eptinezumab (2). Twelve trials focused on episodic migraines, and five trials investigated chronic migraines. The other study recruited populations of both episodic migraine and chronic migraine patients. Moreover, two trials focused on patients who had documented failure to two to four classes of migraine preventive medications. The risk of bias in the studies is shown in Fig. 2. The risk of bias was low in almost all the studies.
Table 1
Characteristics of included studies.
Studies | Trial Registration | Study Design | Migraine Type* | Interventions | Sex (male/female) | Age (mean ± SD) | Course of Treatment |
Bigal 2015a [17] | NCT02021773 | RCT Phase 2b | C | Fremanezumab 900 mg/675-225-225 mg/placebo | 37/227 | 40.7 ± 12.0 | 3 months |
Bigal 2015b [18] | NCT02025556 | RCT phase 2b | E | Fremanezumab 675 mg/225 mg/placebo | 36/261 | 41.2 ± 12.2 | 3 months |
Detke 2018 [19] | NCT02614261 | RCT phase 3 | C | Galcanezumab 240 mg/120 mg/placebo | 167/946 | 41.0 ± 12.1 | 3 months |
Dodick 2014a [20] | NCT01625988 | RCT phase 2 | E | Galcanezumab 150 mg/placebo | 33/184 | 41.4 ± 11.5 | 3 months |
Dodick 2014b [21] | NCT01772524 | RCT phase 2 | E | Eptinezumab 1000 mg/placebo | 30/133 | 38.8 ± 10.2 | 2 months |
Dodick 2018a [22] | NCT02483585 | RCT phase 3 | E | Erenumab 70 mg/placebo | 85/492 | 42.0 ± 11.5 | 3 months |
Dodick 2018b [23] | NCT02629861 | RCT phase 3 | E | Fremanezumab 675 mg single higher dose/225 mg monthly/placebo | 133/742 | 41.8 ± 12.1 | 3 months |
Dodick 2019 [24] | NCT02275117 | RCT phase 2 | C | Eptinezumab 300 mg/100 mg/30 mg/10 mg/placebo | 81/535 | 36.6 ± 9.7 | 3 months |
Ferrari 2019 [25] | NCT03308968 | RCT phase 3b | E/C | Fremanezumab 675 mg quarterly/225 mg monthly/placebo | 138/700 | 46.2 ± 11.1 | 3 months |
Goadsby 2017 [26] | NCT02456740 | RCT phase 3 | E | Erenumab 140 mg/70 mg/placebo | 141/814 | 40.9 ± 11.2 | 6 months |
Reuter 2018 [27] | NCT03096834 | RCT phase 3b | E | Erenumab 140 mg/placebo | 46/200 | 44.4 ± 10.5 | 3 months |
Sakai 2019 [28] | NCT02630459 | RCT phase 2 | E | Erenumab 140 mg/70 mg/28 mg/placebo | 75/400 | 44.0 ± 8.6 | 6 months |
Silberstein 2017 [29] | NCT02621931 | RCT phase 3 | C | Fremanezumab 675 mg quatrerly/675-225-225 mg/placebo | 139/991 | 41.3 ± 12.1 | 3 months |
Skljarevski 2018a [30] | NCT02163993 | RCT phase 2b | E | Galcanezumab 300 mg 120 mg/50 mg/5 mg/placebo | 70/340 | 40.2 ± 12.0 | 3 months |
Skljarevski 2018b [31] | NCT02614196 | RCT phase 3 | E | Galcanezumab 240 mg/120 mg/placebo | 134/781 | 41.8 ± 11.2 | 6 months |
Stauffer 2018 [32] | NCT02614183 | RCT phase 3 | E | Galcanezumab 240 mg/120 mg/placebo | 140/718 | 40.7 ± 11.6 | 6 months |
Sun 2016 [33] | NCT01952574 | RCT phase 2 | E | Erenumab 70 mg/21 mg/7 mg/placebo | 94/389 | 41.3 ± 10.7 | 3 months |
Tepper 2017 [34] | NCT02066415 | RCT phase 2 | C | Erenumab 140 mg/70 mg/placebo | 115/552 | 42.1 ± 11.2 | 3 months |
* C: chronic migraine, E: episodic migraine. |
Publication bias
The probability of publication bias was estimated only by the visual inspection of a funnel plot, since the Review Manager software program, version 5.3, was used. We presented only a funnel plot of the primary outcome, and there was no obvious asymmetry in the funnel plot (Fig. 3) by visual analysis.
MMHDs
The meta-analysis of MMHDs included seventeen trials with a total of 10,007 patients. Obviously, CGRP mAbs were superior in reducing MMHDs compared with placebo for the prevention of migraine (Episodic migraine subgroup, Std. MD -0.42, 95% CI -0.47 to -0.36, I2 = 7%; Chronic migraine subgroup, Std. MD -0.28, 95% CI -0.35 to -0.21, I2 = 13%; Fig. 4). The test for subgroup differences reflected a high heterogeneity among the three subsets (I-square = 91.1%).
50% response rate
Of the 18 studies, seventeen trials with a total of 9061 patients reported data abouts the 50% response rate. CGRP mAbs led to a significantly higher 50% response rate compared with placebo for the prevention of migraines (Episodic migraine subgroup, RR 1.61, 95% CI 1.51 to 1.72, I2 = 47%; Chronic migraine subgroup, RR 1.61, 95% CI 1.41 to 1.83, I2 = 13%; Fig. 5). The subgroup differences were mainly due to the episodic and migraine subsets, which included only Ferrari 2019. In the sensitivity analysis, after the exclusion of Sakai 2019, the RR of the 50% response rate was 1.58 (95% CI: 1.48 to 1.69, I2 = 18%) in the episodic migraines prevented by CGRP mAbs compared with the episodic migraines prevented by placebo.
MHDs and MHHs
Regarding the MHDs, seven trials with a total of 3480 patients reported this outcome. There was a significant difference between CGRP mAbs and placebo in the decrease in MHDs (Episodic migraine: Std. MD -0.31, 95% CI -0.44 to -0.18, I2 = 26%; Chronic migraine: Std. MD -0.28, 95% CI -0.36 to -0.20, I2 = 17%; Fig. 6a). Regarding the MHHs, seven trials with a total of 4346 patients reported this outcome. Similarly, there was a significant difference between CGRP mAbs and placebo in the decrease in MHHs (Episodic migraine: Std. MD -0.34, 95% CI -0.42 to -0.25, I2 = 0%; Chronic migraine: Std. MD -0.23, 95% CI -0.31 to -0.15, I2 = 0%; Fig. 6b).
MSMDs
Seven trials with a total of 3911 patients reported the MSMDs. Using a random -effect model, CGRP mAbs were associated with significantly reduced MSMDs (Std. MD − 0.56, 95% CI − 0.67 to − 0.45, I2 = 60%; Fig. 7) compared with placebo. In the sensitivity analysis, after the exclusion of Dodick 2018a, the same significant result was observed (Std. MD − 0.60, 95% CI − 0.70 to − 0.51, I2 = 33%). Due to the insufficient number of studies, we did not conduct subgroup analysis by type of migraine.
Adverse events
The adverse events were used to evaluate the safety of the drugs. Currently, the CGRP mAbs used to prevent migraine include erenumab, fremanezumab, galcanezumab, and epitinezumab. Therefore, subgroup analysis was used to assess the safety of each CGRP mAb. Fremanezumab was associated with significantly higher incidence of any adverse event when compared with placebo (RR 1.10, 95% CI 1.03 to 1.17, I2 = 0%; Additional file1: S1), and no differences were observed between erenumab and placebo (RR 0.97, 95% CI 0.91 to 1.04, I2 = 39%; Additional file1: S1). Regarding TEAEs, no significant differences were observed between fremanezumab and placebo (RR 1.06, 95% CI 0.89 to 1.27, I2 = 57%; Additional file1: S2), while the incidence of TEAEs after treatment with galcanezumab was significantly higher compared with treatment with placebo (RR 1.11, 95% CI 1.04 to 1.17, I2 = 0%; Additional file1: S2). Unfortunately, the subgroup of fremanezumab included two studies with a high heterogeneity. All the included studies reported withdrawal due to adverse events and SAEs. No significant differences were observed in the withdrawal due to adverse events and SAEs between the four CGRP mAbs and placebo (Additional file1: S3 and S4).
Common adverse events reported in included studies
We analyzed the common adverse events of each CGRP mAb with fixed-effect model. For erenumab, the results showed that the rates of constipation and injection site pain were significantly higher compared to placebo (RR 2.01, 95% CI 1.06 to 3.82, I2 = 31%; RR 1.80, 95% CI 1.10 to 2.94, I2 = 4%; Table 2). For galcanezumab, the meta-analysis showed that the rates of upper respiratory tract infection (RR 1.45, 95% CI 1.03 to 2.05, I2 = 0%), injection site pain (RR 1.31, 95% CI 1.07 to 1.60, I2 = 67%), injection site erythema (RR 2.69, 95% CI 1.63 to 4.45, I2 = 0%), injection site pruritus (RR 15.99, 95% CI 4.47 to 57.24, I2 = 0%), and injection site reaction (RR 4.70, 95% CI 2.68 to 8.25, I2 = 69%) were significantly higher compared with placebo (details are shown in Table 3). There was high heterogeneity in injection site pain and injection site reaction. In the sensitivity analysis, after the exclusion of Skljarevski 2018a, no differences were noted between galcanezumab and placebo in terms of injection site pain (RR 1.22, 95% CI 0.99 to 1.50, I2 = 46%). For injection site reaction, this outcome was still significant after the exclusion of Skljarevski 2018b (RR 3.03, 95% CI 1.66 to 5.50, I2 = 20%). For fremanezumab, the incidence of injection site erythema and injection site induration were significantly higher compared with placebo (RR 1.29, 95% CI 1.06 to 1.57, I2 = 0%; RR 1.27, 95% CI 1.05 to 1.53, I2 = 0%; Table 4). There were only two studies on epitinezumab in the included studies, and we analyzed the adverse events reported jointly by these two studies. The results showed that epitinezumab led to rates of dizziness, nausea, migraine, and upper respiratory tract infection that were similar to those caused by placebo (Table 5).
Table 2
Common adverse events of erenumab and placebo.
Common adverse events | Number of studies | Erenumab | Placebo | RR | P | I2 |
Nausea | 4 | 33/1400 | 28/1043 | 0.91 [0.55, 1.52] | 0.73 | 0% |
Fatigue | 4 | 30/1141 | 19/885 | 1.27 [0.73, 2.22] | 0.41 | 0% |
Constipation | 4 | 41/1566 | 13/1026 | 2.01 [1.06, 3.82] | 0.03 | 31% |
Back pain | 4 | 26/1130 | 15/732 | 1.14 [0.60, 2.18] | 0.69 | 0% |
Arthralgia | 2 | 15/739 | 11/472 | 0.88 [0.38, 2.02] | 0.75 | 29% |
Migraine | 4 | 24/1400 | 23/1043 | 0.80 [0.46, 1.39] | 0.44 | 49% |
Nasopharyngitis | 6 | 185/1791 | 129/1303 | 0.89 [0.72, 1.10] | 0.27 | 30% |
Sinusitis | 2 | 24/916 | 13/608 | 1.19 [0.60, 2.35] | 0.62 | 0% |
Upper respiratory tract infection | 5 | 71/1519 | 39/1167 | 1.30 [0.89, 1.90] | 0.18 | 0% |
Influenza | 3 | 24/1022 | 21/761 | 0.93 [0.51, 1.68] | 0.8 | 0% |
Injection site pain | 5 | 51/1751 | 24/1150 | 1.80 [1.10, 2.94] | 0.02 | 4% |
Table 3
Common adverse events of galcanezumab and placebo.
Common adverse events | Number of studies | Galcanezumab | Placebo | RR | P | I2 |
Dizziness | 3 | 31/987 | 24/1003 | 1.31 [0.78, 2.22] | 0.31 | 0% |
Nausea | 3 | 21/670 | 29/679 | 0.74 [0.42, 1.28] | 0.27 | 0% |
Fatigue | 2 | 23/1009 | 22/1019 | 1.06 [0.59, 1.88] | 0.85 | 0% |
Diarrhoea | 2 | 19/1009 | 20/1019 | 0.96 [0.52, 1.79] | 0.9 | 0% |
Back pain | 3 | 30/1088 | 28/1100 | 1.09 [0.65, 1.80] | 0.75 | 16% |
Neck pain | 3 | 18/1088 | 14/1100 | 1.30 [0.65, 2.60] | 0.46 | 0% |
Abdominal pain | 2 | 16/662 | 12/668 | 1.35 [0.64, 2.83] | 0.43 | 0% |
Arthralgia | 2 | 12/662 | 12/668 | 1.02 [0.46, 2.23] | 0.97 | 0% |
Migraine | 2 | 16/981 | 9/990 | 1.79 [0.80, 4.04] | 0.16 | 0% |
Oropharyngeal pain | 2 | 14/981 | 6/990 | 2.36 [0.91, 6.11] | 0.08 | 0% |
Nasopharyngitis | 5 | 95/1679 | 105/1698 | 0.92 [0.70, 1.20] | 0.52 | 0% |
Sinusitis | 3 | 33/1088 | 23/1100 | 1.45 [0.86, 2.46] | 0.16 | 15% |
Upper respiratory tract infection | 4 | 73/1253 | 51/1266 | 1.45 [1.03, 2.05] | 0.04 | 0% |
Influenza | 3 | 31/1435 | 22/1451 | 1.43 [0.83, 2.45] | 0.2 | 25% |
Injection site pain | 5 | 193/1679 | 149/1698 | 1.31 [1.07, 1.60] | 0.008 | 67% |
Injection site erythema | 4 | 54/1542 | 20/1561 | 2.69 [1.63, 4.45] | 0.0001 | 0% |
Injection site pruritus | 3 | 39/1435 | 2/1451 | 15.99 [4.47, 57.24] | ༜0.0001 | 0% |
Injection site reaction | 3 | 67/1435 | 14/1451 | 4.70 [2.68, 8.25] | ༜0.00001 | 69% |
Table 4
Common adverse events of fremanezumab and placebo.
Common adverse events | Number of studies | Fremanezumab | Placebo | RR | P | I2 |
Dizziness | 3 | 35/1508 | 8/756 | 2.10 [1.00, 4.40] | 0.05 | 0% |
Nausea | 4 | 28/2089 | 26/1049 | 0.54 [0.32, 0.92] | 0.02 | 16% |
Fatigue | 3 | 31/1334 | 8/674 | 1.95 [0.90, 4.23] | 0.09 | 0% |
Constipation | 2 | 11/735 | 2/366 | 2.30 [0.59, 8.96] | 0.23 | 0% |
Insomnia | 2 | 16/735 | 2/366 | 3.29 [0.87, 12.35] | 0.08 | 0% |
Paraesthesia | 2 | 5/366 | 3/193 | 0.83 [0.22, 3.17] | 0.78 | 68% |
Back pain | 3 | 15/927 | 8/470 | 0.96 [0.41, 2.24] | 0.92 | 0% |
Neck pain | 2 | 7/735 | 0/366 | 4.02 [0.50, 32.06] | 0.19 | 0% |
Arthralgia | 2 | 7/753 | 3/381 | 1.10 [0.31, 3.84] | 0.89 | 12% |
Migraine | 2 | 4/366 | 2/193 | 0.95 [0.20, 4.46] | 0.95 | 0% |
Headache | 2 | 7/366 | 2/193 | 1.58 [0.38, 6.61] | 0.53 | 48% |
Nasopharyngitis | 5 | 82/2263 | 46/1138 | 0.89 [0.63, 1.27] | 0.53 | 0% |
Sinusitis | 4 | 29/1702 | 22/861 | 0.67 [0.39, 1.16] | 0.15 | 0% |
Upper respiratory tract infection | 4 | 80/2089 | 37/1049 | 1.09 [0.74, 1.59] | 0.67 | 0% |
Depression | 2 | 2/366 | 3/193 | 0.39 [0.09, 1.78] | 0.23 | 62% |
Influenza | 2 | 11/753 | 3/381 | 1.86 [0.52, 6.64] | 0.34 | 0% |
Bronchitis | 2 | 15/773 | 3/397 | 2.28 [0.72, 7.21] | 0.16 | 0% |
Injection site pain | 5 | 433/2263 | 197/1138 | 1.10 [0.95, 1.27] | 0.19 | 0% |
Injection site erythema | 5 | 308/2263 | 119/1138 | 1.29 [1.06, 1.57] | 0.01 | 0% |
Injection site pruritus | 3 | 16/927 | 3/470 | 2.15 [0.73, 6.35] | 0.16 | 0% |
Injection site bruising | 3 | 15/927 | 5/470 | 1.54 [0.56, 4.20] | 0.4 | 0% |
Injection site induration | 3 | 317/1897 | 125/945 | 1.27 [1.05, 1.53] | 0.01 | 0% |
Injection site hemorrhage | 2 | 27/1336 | 16/668 | 0.84 [0.46, 1.55] | 0.59 | 0% |
Table 5
Common adverse events of eptinezumab and placebo.
Common adverse events | Number of studies | Eptinezumab | Placebo | RR | P | I2 |
Dizziness | 2 | 17/324 | 10/203 | 0.95 [0.45, 1.99] | 0.89 | 22% |
Nausea | 2 | 20/324 | 11/203 | 1.02 [0.50, 2.08] | 0.95 | 0% |
Migraine | 2 | 9/324 | 4/203 | 1.36 [0.40, 4.66] | 0.63 | 0% |
Upper respiratory tract infection | 2 | 28/324 | 12/203 | 1.50 [0.77, 2.94] | 0.23 | 0% |